Rodents cause significant damage to lowland rice crops across Cambodia and farmers are increasingly turning to chemical rodenticides for pest rodent management. Concern about the potential damage of these chemicals to human and environmental health provides the impetus for a search for alternative rodent management strategies, with particular interest in ecologically-based rodent management (EBRM). EBRM has a strong emphasis on non-chemical methods including the Trap Barrier System (TBS) and uses knowledge of pest rodent ecology to design effective interventions that limit population growth. For Cambodia, almost nothing was known at the start of this study regarding the identity, distribution and ecology of the country’s rodent pests. This research was designed to document this knowledge in parallel with a rodent management trial carried out at Somrong Commune in Kampong Cham Province of Cambodia, with the practical aim of informing on the outcomes of the trial and assisting with future development of EBRM in Cambodia. The large physical scale of the study and the relatively short time frame available for research led me to develop an action research approach that combined traditional positivist biological research with a more constructivist approach and participatory methods to gain access to and utilize local knowledge about rats. This research documents Cambodian pest rodents in lowland rice field cropping systems, through a combination of nation-wide collecting followed by detailed taxonomic assessments, and the use of farmer meetings and interviews to estimate the history and severity of the problem in each area. A total of nine rodent pest species were identified but both the pest rodent community and the severity of the associated problems vary from province to province. One species, the rice field rat (Rattus argentiventer), appears to be actively spreading and is not yet found in the northwestern provinces. A more detailed study in Somrong Commune found seven species and documented aspects of habitat use, breeding biology and movement. The inclusion of local knowledge in the study added a spatial and temporal scale to the study that could not have been obtained through conventional means, and which led to novel hypotheses about pest rodent ecology in the Somrong landscape, presented in the form of a heuristic landscape model. Criteria are developed for assessing the rigour and reliability or trustworthiness of the results. Particular attention is paid to the potential value of ‘expert’ knowledge which is rarely used in participatory research but which holds enormous potential for research into technical issues. The results of the TBS trial carried out in Somrong Commune were analysed to assess whether implementation of this method resulted in lower levels of rodent damage and higher rice yields relative to a single control commune, Lvea, and whether the farming community in Somrong is likely to adopt the method in the future. Although some farmer survey data suggest that yields improved in Somrong during the TBS trial, statistical analysis of the quantitative data set fails to yield compelling evidence of any benefit from use of TBS. The results emphasize the important role of landscape factors, especially the annual flooding cycle of the Mekong River, in determining the spatial and temporal distribution of rodent damage. Analysis of the relationship between rat captures and claimed yield increases also suggests that the TBS was not primarily responsible for the yield increase—too few rats were captured to account for the additional yield, even under extreme scenarios for how individual rodents cause damage to rice crops. Somrong farmers are unlikely to continue with the TBS method after completion of the project due to many perceived technical problems with implementing the method, and other concerns over its high labour and monetary costs. The ecological results of this study are used to suggest various alternative means by which Somrong farmers might combat the problem of rodent pests. Finally, a new approach is recommended for gaining an appreciation of rodent ecology on a landscape scale. This approach uses a combination of generalities about rodent biology and ecology, field observations, and incorporates local knowledge from local ‘experts’. This approach is likely to yield a faster and more complete understanding of the spatial and temporal dimensions of a local rodent community than more conventional approaches and will hopefully lead to more effective and relevant rodent management recommendations.

In Cambodia, over 80% of the population is engaged in agricultural production, mostly in rice cultivation. The country, however, faces a lot of problems of rice crop damage. In Cambodia, rodents are considered to be one of the most important pests of rice, with the potential to cause severe damage during periodic plague events. As is seen in other developing countries, physical methods of rodent control are probably the most commonly used approaches. In general, the choice of methods varies with the availability of resources, season, the participation of farmers, the condition of the rice field, weather conditions, experience with the success of the techniques, likely level of rodent attack and other considerations. However, effective rodent control methods suiting farmers’ needs and applicable to the agricultural production environment in Cambodia are yet to be assessed. The aim of this research was to examine the socio-economic factors impacting on the implementation of several rodent control methods in comparison to the Trap-Barrier System (TBS) in order to improve rodent pest control in Cambodia. The research employed a constructivist approach in which the researcher played a role as a participant observer, to study the social setting and realities as constructed by the participants in the study sites. Rodent management was studied as a complex phenomenon, occurring within the milieu of family, village and commune life in rural Cambodia. The study looks into the beliefs and behaviours of the farming community and examines how Cambodian traditions affect the implementation of rodent management techniques and the management of rodents as a community activity. Participatory techniques were employed to gather information concerning the farmers’ assessments of the effectiveness of various rodent management options practiced in Samrong Commune. The level of adoption and adaptation of TBS in the commune was observed over a period of several years. This study is the first detailed study of farmer perceptions of alternative rodent control methods in Cambodia, especially the TBS. The study was based on long term trials involving a TBS application on a commune scale and investigations after the removal of all subsidies for purchase of materials. It was apparent that TBS use declined and eventually become limited to a few individuals who had worked closely with the project. The study also demonstrates how BBNs and Netica software can be applied as participatory tools to develop and explore the decision making structure of farmers. These tools can also be used as extension tools and can contribute to better decision making by communities. The study also shows how an understanding of traditional aspects of public goods management by communities can contribute to the design of effective contemporary solutions to problems involving the community and public goods management. In this sense its significance goes well beyond the specific context of rodent control, pest management and rice production and takes on wider significance in the field of community development generally. The study applied a mixed qualitative and quantitative approach to data collection and analysis, involving interviews, surveys, modelling, workshops, field notes, observations and document review. Five rodent control methods (TBS, rodenticide, electric fencing, netting and hunting) were found to be the most common rodent control methods used in the study sites. Some common factors were found to influence the effectiveness of these methods. Farmer participation was found to be the most widely influential factor affecting the rodent management techniques. Netting method was considered the most effective method of rodent control by farmers, as well as the method that had the best benefit-cost ratio. This technique was only found to be applicable in the wet season. In general, the effectiveness of TBS and its benefit-cost ratio was ranked fourth of the five methods, because it required high farmer participation, monitoring and maintenance as well as high material expense. Improvement of TBS can be made when it can be clearly demonstrated to farmers that TBS is an applicable method, being safe and able to catch more rats. To suit the farmers’ requirements for an effective rodent control method, TBS must be able to deal with migrating rats, situations of high rat abundance and the planting problems associated with an irregular rain regime. To achieve these conditions, some intervention policies are required in terms of research and extension support.

MENVSC / Department of Ecology and Resource Management / Rodents can quickly respond to land use changes whether the change positively or negatively influences their life. In the case of positive influence, rodents exploit the additional food resources and increase their numbers to potential pest level, especially in the absence of predators. Such a population increase can potentially be harmful to humans due to the diseases that rodents carry and the costs due to damage to crops, stored foods and personal possessions that they may cause to small holder farmers. Small holder farmers live in a mixed landscape that is constantly changing. Such changes are changes in land use and they do not only affect rodent population dynamics and species composition, but also their ecosystem services and integrity. Understanding how rodents respond to these land use changes (crop, grazing and settlement) will not only improve the implementation of Ecologically Based Rodent Management (EBRM), but might also enable the monitoring of ecosystem integrity. Rodent trapping was conducted in two different study sites which experience different rainfalls during wet and dry season. A 70 m x 70 m grid was set in three different land uses (crops, grazing and settlement) per study site. A mark-recapture technique was applied and all captures were processed on a temporal station on site. In all grids at a distance of 30 m a line of 20 snap traps were set. With the tapping effort of 1470 trap nights per season for both seasons in this study we captured 839 rodents and 2 shrews, which represented 469 individual rodents and 1 individual shrew. At Vyeboom, cropping land use had the highest number capture (210) and the highest in species richness (9) rodent species and 1 shrew. The settlement land use was second at 144 captures for 8 rodent species and lastly the grazing land use at 80 captures with 7 rodent species. On the other hand, at Ka-Ndengeza also cropping land use had the highest capture (186) with highest richness at 7 rodent species. When it comes to settlement and grazing, settlement was second (129) to cropping in terms of the number of capture but last in terms of richness (5) whilst grazing was last in terms of number of captures 92 and second in terms of richness (6). Despite the high diversity of rodents, only Mastomys natalensis, Gerbilliscus leucogaster, Steatomys pratensis and Rattus rattus were captured in meaningful sample sizes to allow for robust density estimation. Similarly there were strong seasonal effects on rodent captures, with almost no captures during the wet season.